CN109784996A - A kind of personalized distribution method and system of favor information - Google Patents
A kind of personalized distribution method and system of favor information Download PDFInfo
- Publication number
- CN109784996A CN109784996A CN201910023753.7A CN201910023753A CN109784996A CN 109784996 A CN109784996 A CN 109784996A CN 201910023753 A CN201910023753 A CN 201910023753A CN 109784996 A CN109784996 A CN 109784996A
- Authority
- CN
- China
- Prior art keywords
- preferential
- information
- action message
- favor information
- movable
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
Landscapes
- Information Retrieval, Db Structures And Fs Structures Therefor (AREA)
- Management, Administration, Business Operations System, And Electronic Commerce (AREA)
Abstract
The invention discloses the personalized distribution methods and system of a kind of favor information, are related to personal marketing field, comprising: when getting movable membership information, according to member's feature in default basic data and movable membership information, matching obtains corresponding member；The corresponding favor information of the activity membership information is distributed into the member.The present invention can match suitable favor information by its history consumer behavior before member (i.e. customer) does not consume by big data technology for its amount body, promote its consumption；And distribution favor information is obtained according to default basic data multi dimensional analysis, the subjective factor influence of marketing plan personnel is lower, and matching precision is high, substantially increases the usage experience of member.
Description
Technical field
The present invention relates to personal marketing field more particularly to the personalized distribution methods and system of a kind of favor information.
Background technique
Retailer understands not timing and carries out marketing activity, promotes the promotion of sale, and traditional means are using stereotyped, high
The public marketing activity of universality, extensively distributes leaflets.Such as: December 11, apple is reduced to 3.8/ jin by 4.8/ jin.This marketing is lived
Dynamic precision is lower, and the promotion effect reached is unsatisfactory.
For this problem, it is known that some marketing system service providers by being got through with retailer's cash register system after, pass through
The purchaser record actually generated after consumer's clearing triggers corresponding marketing activity, so that precision marketing is realized in part.Such as: purchase
The member that customer's unit price reaches 300 gives the whole audience 8 folding certificate, buys objective unit price and reaches 200 and give the whole audience 9 folding certificate, buys visitor single
Valence reaches 100 and gives 9.5 folding certificate of the whole audience etc..
But above-mentioned this mode mainly has following problems: in the activity discount coupon distribute be occur customer consumption it
Afterwards, it is difficult that customer is allowed to carry out second of consumption at once, causes movable ineffective；And in the activity discount coupon trigger condition
It is business experience, subjective specified service logic of the marketing plan personnel according to oneself, inadequate scientific and precise.
Summary of the invention
The object of the present invention is to provide the personalized distribution methods and system of a kind of favor information, measure status for each customer
With favor information, precision and animation effect are improved.
Technical solution provided by the invention is as follows:
A kind of personalized distribution method of favor information, comprising: when getting movable membership information, according to default basis
Member's feature in data and the movable membership information, matching obtain corresponding member；The movable membership information is corresponding
Favor information distribute to the member.
In the above-mentioned technical solutions, favor information, this preferential letter are distributed to it by delineation member's characteristic matching member
The method of salary distribution of breath is suitble to marketing demand to use when having clear member group, accurately distributes to particular member, has high matching
Precision.And the method for salary distribution can be matched properly by its history consumer behavior for its amount body before member (i.e. customer) does not consume
Favor information, promote its consumption.
Further, further includes: when the quantity for the member that matching obtains is less than the preferential quantity of the movable membership information,
The member is expanded using crowd's broadcast algorithm, the quantity of the member is made to reach the preferential quantity.
In the above-mentioned technical solutions, the expansion of member can screen the higher member of matching degree, guarantee matching
Precision, meanwhile, the favor information of the preferential quantity of setting can also be dispensed.
Further, when getting allocation strategy, at least one preferential action message and at least one movable membership information,
According to default basic data, member corresponding with each preferential action message is filtered out respectively；Respectively by each preferential work
The dynamic corresponding member of information is ranked up by the active policy in the preferential action message, obtains each preferential action message
Corresponding degree of correlation list；According to member's feature in default basic data and each movable membership information, matching obtains each
The corresponding member of the activity membership information；According to each movable corresponding member of membership information and each preferential activity letter
Corresponding degree of correlation list is ceased, is each member point in default basic data according to the pre-set priority in the allocation strategy
Favor information with preset quantity in the allocation strategy.
In the above-mentioned technical solutions, two kinds of methods of salary distribution are used in combination, and while meeting different marketing purposes, ensure that
Matching precision improves the effect of marketing activity.
Further, further includes: when the quantity for the favor information being assigned there are a member is not up to the preset quantity,
It automatically is the remaining favor information of member's polishing.
In the above-mentioned technical solutions, if can not consume industry according to the history of member distributes appropriate number of favor information,
Him can be given polishing at random or regularly from remaining favor information, complete the distribution of favor information.
Further, the basis presets basic data, filters out member corresponding with the preferential action message and includes:
According to default basic data, the specific quartile of each member visitor unit price in preset time period is calculated in default basic data
Number；The specific quantile is corresponding with the active policy in the preferential action message；According to the specific quantile and excellent
The amount of money after favour filters out corresponding member according to the active policy；Wherein, the preferential rear amount of money is according to the preferential work
What the type of offer in dynamic information was calculated.
In the above scheme, different members is matched, using different quantiles to meet different active policy needs
Member improves the matching precision of member and favor information.
The present invention also provides a kind of personalized distribution systems of favor information, comprising: obtains module, acquisition activity member letter
Breath；Matching module, for according to the member's feature preset in basic data and the movable membership information, matching to obtain corresponding
Member；Distribution module, for the corresponding favor information of the activity membership information to be distributed to the member.
Further, further includes: extension module, the quantity of the member for obtaining when matching are less than the movable membership information
In preferential quantity when, the target members are expanded using crowd's broadcast algorithm, the quantity of the member is made to reach the activity
Preferential quantity in membership information.
Further, the acquisition module is further used for obtaining allocation strategy, at least one preferential action message and at least
One movable membership information；The personalized distribution system of the favor information, further includes: screening module, for according to default
Basic data filters out member corresponding with each preferential action message respectively；Sorting module, being used for respectively will be each described excellent
The corresponding member of favour action message is ranked up by the active policy in the preferential action message, obtains each preferential activity
The corresponding degree of correlation list of information；The matching module is further used for according to default basic data and each movable member
Member's feature in information, matching obtain the corresponding member of each activity membership information；The distribution module, is further used for
According to the corresponding member of each movable membership information and the corresponding degree of correlation list of each preferential action message, according to described
Pre-set priority in allocation strategy is that each member in default basic data distributes preset quantity in the allocation strategy
Favor information.
Further, further includes: polishing module, for being not up to institute when the quantity for the favor information being assigned there are a member
It is automatically the remaining favor information of member's polishing when stating preset quantity.
Further, the screening module, for filtering out and a preferential action message pair according to basic data is preset
The member answered includes: screening module, according to default basic data, is calculated in default basic data in preset time period each
The specific quantile of member's visitor's unit price；The specific quantile is corresponding with the active policy in the preferential action message；With
And according to the specific quantile and the preferential rear amount of money, corresponding member is filtered out according to the active policy；Wherein, described
The amount of money is calculated according to the type of offer in the preferential action message after preferential.
Compared with prior art, the personalized distribution method of favor information of the invention and system beneficial effect are:
The present invention can be it by its history consumer behavior before member (i.e. customer) does not consume by big data technology
It measures body and matches suitable favor information, promote its consumption；And distribution favor information is according to default basic data multi dimensional analysis
It obtains, the subjective factor influence of marketing plan personnel is lower, and matching precision is high, substantially increases the usage experience of member.
Detailed description of the invention
Below by clearly understandable mode, preferred embodiment is described with reference to the drawings, to a kind of individual character of favor information
Above-mentioned characteristic, technical characteristic, advantage and its implementation for changing distribution method and system are further described.
Fig. 1 is the flow chart of personalized distribution method one embodiment of favor information of the present invention；
Fig. 2 is the flow chart of the variant embodiment of Fig. 1；
Fig. 3 is the schematic diagram of one embodiment of movable membership information distribution；
Fig. 4 is the flow chart of personalized another embodiment of distribution method of favor information of the present invention；
Fig. 5 is the structural schematic diagram of personalized distribution system one embodiment of favor information of the present invention；
Fig. 6 is the structural schematic diagram of the variant embodiment of Fig. 5；
Fig. 7 is the structural schematic diagram of personalized another embodiment of distribution system of favor information of the present invention.
Drawing reference numeral explanation:
10. obtaining module, 20. screening modules, 30. sorting modules, 40. distribution modules, 50. matching modules, 60. expand mould
Block, 70. polishing modules.
Specific embodiment
In order to more clearly explain the embodiment of the invention or the technical proposal in the existing technology, Detailed description of the invention will be compareed below
A specific embodiment of the invention.It should be evident that drawings in the following description are only some embodiments of the invention, for
For those of ordinary skill in the art, without creative efforts, it can also be obtained according to these attached drawings other
Attached drawing, and obtain other embodiments.
To make simplified form, part related to the present invention is only schematically shown in each figure, they are not represented
Its practical structures as product.In addition, there is identical structure or function in some figures so that simplified form is easy to understand
Component only symbolically depicts one of those, or has only marked one of those.Herein, "one" is not only indicated
" only this ", can also indicate the situation of " more than one ".
The personalized distribution method and system of favor information of the invention are realized based on big data technology.
In one embodiment of the invention, as shown in Figure 1, a kind of personalized distribution method of favor information, comprising:
S101 is special according to the member in default basic data and movable membership information when getting movable membership information
Sign, matching obtain corresponding member；
By corresponding (type of offer the is corresponding) favor information of movable membership information, (favor information is in S102 in different forms
It is existing, such as: discount coupon, red packet information, bonus information etc.) distribute to and (be matched to) member.
Specifically, can be corresponding by member's feature of delineation member when wanting to distribute favor information for particular member
Member distributes favor information.
Movable membership information includes: type of offer, preferential quantity and member's feature, these three all can be clever according to actual needs
Setting living.
Type of offer includes:
1) whole audience favor information, such as: the whole audience expires 100-99, and the whole audience expires 500-230, and the whole audience full 100 send 20 yuan of red packets, entirely
Send 10 yuan of bonuses etc. in field full 200.
2) category favor information, category refer to the belonging kinds of each commodity, such as: the category of Coca-Cola is carbonic acid drink
Material.Category favor information refers to the activity done in this category, such as: dilated food class, full 160-10, full 56 send 5 yuan it is red
Packet, full 78 send 10 yuan of bonuses etc..
3) brand favor information, brand refer to some brand, such as: first-class good, Maxam brand etc..Brand favor information is
Refer to the activity done in this brand, such as: pleasure expires 50-10, and Chef Kang full 39 additionally send packet instant noodles etc..
4) single-item favor information, single-item refer to some commodity, and only this commodity does activity, such as: hundred million grow Ao Liao folder
Heart biscuit original flavor 349g expires 50-5；Happy treasured drinking water pure water 4.5L*4 branch/case buys 2 casees happy precious drinking water for sending 1 bottle of 500ml
Deng.
Preferential quantity refers to the quantity of favor information, such as: the whole audience expires 500-230 and there was only 3000.
Member's feature includes: the label of member, specified consumer behavior etc..
Default basic data be to data source (such as: retailer do shopping receipt transaction data, information of goods information data, member
Information data and shops's information data) import (such as: uncorrected data load, data cleansing, data accuracy inspection, data map,
Standardization number) after, corresponding label is stamped for each member according to these data, and, the sale sorted out summarizes data,
It is used for the subsequent corresponding member of screening.It can be implemented by using the prior art for the mode that member labels, details are not described herein.
Each member has multiple member's labels, defines according to actual needs.Such as: the label of loyalty is played to member, with
Lower 4 select 1:a) loyal customer, b) chance customer, c) be easy to run off client, d) new client, according to the historical information of member's shopping to every
A member stamps corresponding loyalty label.For another example: gender；Age；Price attention rate, a) quality pays close attention to customer, b) mainstream
Customer, c) price pay close attention to client；Life style, a) cook, b) household purchasing, c) have that child, d) promotion is sensitive, e) health, f) new
Product；Preference shops etc..
Sale summarizes data and determines specific dimension according to actual needs (such as: the data that member's feature of definition needs)
Data form, such as: a. product x crowd label x shops, 4 weeks customer number summary sheets；B. shops x crowd label, 4 Zhous customers
Number summary sheet；C. people x brand, 52 weeks purchase number summary sheets；D. people x category, 52 weeks purchase number summary sheets；E. each meeting
52 weeks shopping number summary sheets of member.
The example for matching corresponding member from default basic database by member's feature is as follows: as shown in figure 3, setting
Member's feature are as follows: have and be easy to run off member, chance member, loyal member's label, and buy overnutrition health care product in one month
Member, corresponding member is filtered out from default basic data by these member's features.
Optionally, as shown in Figure 2, further includes: S203 is less than movable membership information when the quantity for the member that matching obtains
When preferential quantity, member is expanded using crowd's broadcast algorithm, the quantity of member is made to reach preferential quantity.By remaining favor information
Distribute to the member found using crowd's broadcast algorithm.
Specifically, crowd's broadcast algorithm can be used: Look-alike algorithm, certainly, the calculation of other achievable member's extensions
Method is also suitable, and with no restriction.
Crowd's broadcast algorithm is to find from remaining member and the assigned the most similar member of member, fortune
Logic when row includes:
1, consider feature: the score of all labels for the member being assigned, type of offer corresponding brand and category.Product
The score rule of board and category can be defined flexibly, such as: the accumulative gold of this brand, category is bought in each assigned member 52 weeks
Volume is as score.
2, it calculates and is assigned population centers' point, such as: the average value of all features.
3, member and assigned people of each preference shops within the scope of the corresponding movable shops of movable membership information are calculated
The distance of group center's point.
4, by widened member needed for apart from ascending sequence, selecting.
In the present embodiment, favor information, point of this favor information are distributed to it by delineation member's characteristic matching member
It is suitble to marketing demand to use when having clear member group with mode, accurately distributes to particular member, there is high matching precision.
In another embodiment of the present invention, as shown in figure 4, a kind of personalized distribution method of favor information, comprising:
S301 obtains allocation strategy, at least one preferential action message and at least one movable membership information；
S302 filters out member corresponding with each preferential action message according to default basic data respectively；
S303 is respectively by the corresponding member of each preferential action message by active policy (the corresponding row in preferential action message
Sequence strategy) it is ranked up, obtain the corresponding degree of correlation list of each preferential action message；Active policy is different, in degree of correlation list
The ordering strategy of each member can be different, also different, determine according to the actual situation.
For S304 according to member's feature in default basic data and each movable membership information, matching obtains each movable member's letter
Cease corresponding member；The sequence of S302-303 and S304 is not construed as limiting, can be parallel, can also S304 preceding, S302-S303 is rear
Deng.
S305 according to each movable corresponding member of membership information and the corresponding degree of correlation list of each preferential action message, according to
Pre-set priority in allocation strategy is the preferential of preset quantity in each member distribution allocation strategy preset in basic data
Information (favor information is presented in different forms, such as: discount coupon, red packet information, bonus information etc.).
Specifically, preferential action message includes: type of offer, preferential quantity, active policy.Type of offer, preferential quantity
With active policy flexible setting according to actual needs.The definition of type of offer and preferential quantity is same as the previously described embodiments, herein
It repeats no more.
There are many active policies, such as: brand feedback strategy, single-item consumption escalation policy, the whole audience mention objective single strategy, brand
New strategy, category solicit patrons Flow Policy, whole audience solicit patrons Flow Policy etc. are drawn, it is associated with type of offer.
In preferential action message, according to different type of offer, different active policies is set, improves sale to reach
The different marketing purposes such as amount, raising member's amount.
Default basic data is same as the previously described embodiments, and details are not described herein, it should be noted that the present embodiment to be adopted
Therefore the mode distributed with preferential action message calculates sale and summarizes data other than considering member's feature, it is also necessary to consider excellent
The needs of favour action message.
The member that each preferential action message is directed to is different, and screening process naturally also has difference, according to preferential activity
Active policy screening in information.
Optionally, according to default basic data, filtering out member corresponding with a preferential action message includes a variety of sides
Formula determines according to the active policy of actual set, names some examples for reference:
The first, is according to default basic data, be calculated in default basic data in preset time period (such as: 52 weeks)
The specific quantile of each member visitor unit price；Specific quantile is corresponding with the active policy in preferential action message；According to spy
Determine quantile and the preferential rear amount of money, filters out corresponding member according to active policy；Wherein, the preferential rear amount of money is according to preferential work
What the type of offer in dynamic information was calculated.
Such as: type of offer is that the whole audience expires 399-100, and active policy is whole audience solicit patrons Flow Policy, screens corresponding member
Process are as follows: 1) visitor's monovalent (i.e. each shopping list) in 52 weeks is pass by according to each member and calculates each 25 quantile of member；2)
Monovalent 25 quantiles of 52 weeks visitors of filtering out over are greater than the member of the preferential rear amount of money, and the preferential rear amount of money here is 399-100=
299。
Type of offer is that the whole audience expires 199-50, and active policy is that the whole audience mentions objective single strategy, screens the process of corresponding member
Are as follows: 1) visitor's monovalent (i.e. each shopping list) in 52 weeks is pass by according to each member and calculates each 50 quantile of member；2) it filters out
Monovalent 50 quantiles of past 52 weeks visitor are greater than the preferential rear amount of money and the preferential rear amount of money is less than the monovalent meeting of highest visitor in 52 weeks in the past
Member, the preferential rear amount of money here is 199-50=149.
Type of offer is that a certain category expires 59-10, and active policy is category solicit patrons Flow Policy, screens the mistake of corresponding member
Journey are as follows: 1) visitor's monovalent (i.e. each shopping list) in 52 weeks is pass by according to each member and calculate each 25 quantile of member；2) it screens
Pass by the member that monovalent 25 quantiles of 52 weeks visitors are greater than the preferential rear amount of money out, the preferential rear amount of money here is 59-10=49.
Type of offer is that a certain category expires 59-10, and active policy is that category mentions objective single strategy, screens the mistake of corresponding member
Journey are as follows: 1) visitor's monovalent (i.e. each shopping list) in 52 weeks is pass by according to each member and calculate each 50 quantile of member；2) it screens
Pass by monovalent 50 quantiles of 52 weeks visitors out greater than the preferential rear amount of money and the preferential rear amount of money is less than the objective monovalent meeting of highest in 52 weeks in the past
Member, the preferential rear amount of money here is 59-10=49.
Different active policies selects different quantile and screening conditions, guarantees corresponding type of offer and screens
Member more match.Such as: mentioning visitor is singly to measure in order to which the visitor for improving purchase is single, and 50 quantiles are greater than the preferential rear amount of money, explanation
The both greater than preferential rear amount of money of visitor's unit price that member's half is bought, the preferential rear amount of money are less than the 52 weeks in the past monovalent members of highest visitor,
Illustrate that this kind of member also has an opportunity to buy the preferential rear amount of money with ability, the member for meeting the two conditions, which belongs to, can potentially mention
Objective single member.The corresponding member of preferential action message is filtered out according to the history consumer behavior of member, matching degree is higher, for meeting
Member receives the favor information oneself made to measure and lays the foundation.
Second, according to the type of offer in preferential action message, calculate the corresponding reference price band of type of offer；It calculates
The corresponding category price zone of type of offer of each member in default basic data；According to the product of reference price band and each member
Class price zone filters out corresponding member according to the active policy in preferential action message.
Such as: type of offer is that brand A expires 49-5, and active policy is brand consumption escalation policy, screens corresponding member
Process are as follows: 1, calculate the corresponding reference price band of brand A and (divide price zone 1-5 grades, 1 grade for the brand A category A being related to
It is minimum), reference price band is 3 grades；2, data are summarized according to the sale preset in basic data, calculates each membership buying category
Price zone locating for the flat fare of each single-item under A, i.e., the corresponding category price zone of the type of offer of each member；Filter out category
Price zone is lower than all members of reference price band, i.e. category price zone is located at 1 or 2 grade of member.
Type of offer is that single-item B full 39 is sent one bottle of mineral water (cost price is 3 yuan), and active policy is single-item consumption upgrading plan
Slightly, the process of corresponding member is screened are as follows: 1, the corresponding reference price band of calculating single-item B (category B strokes be related to for single-item B
Divide price zone 1-5 grades, 1 grade is minimum), reference price band is 4 grades；2, data are summarized according to the sale preset in basic data, counted
Calculate price zone locating for the flat fare of each single-item under each membership buying category B, i.e., the corresponding product of type of offer of each member
Class price zone；All members that category price zone is lower than reference price band are filtered out, i.e., category price zone is positioned at 1 or 2 or 3 grade
Member.
The third calculates people's certificate of all members according to the type of offer in default basic data and preferential action message
The degree of correlation；According to the active policy in people's certificate degree of correlation and preferential action message, corresponding member is filtered out.
Such as:
Type of offer is to send the red packet of 5 yuan of brand C, and active policy is brand feedback strategy, screens the mistake of corresponding member
Journey are as follows: 1, according to preset basic data in each member give a mark to the history consumer behavior of each single-item under brand C, give a mark system
Flexible setting, such as: the accumulative amount of each single-item added up 256 yuan, and was just 256 points as score value under brand C in past 52 weeks；
2, brand feedback strategy is the member filtered out greater than 500.
Type of offer is to send 1 single-item D, and active policy is brand feedback strategy, screens the process of corresponding member are as follows: 1,
It is given a mark according to member each in default basic data to the history consumer behavior of single-item D, according to cumulative consumption, calculates flat fare
Lattice are as score；2, brand feedback strategy is the member filtered out greater than 100.
4th kind, according to the active policy in default basic data and preferential action message, filter out do not bought it is preferential
The member of the corresponding commodity of type of offer in action message.
Such as:
Type of offer is that single-item E expires 29-10, and active policy is that single-item draws new strategy, screens the process of corresponding member are as follows:
Filter out all members for not buying single-item E.
Type of offer is that brand F expires 59-20, and active policy is that brand draws new strategy, screens the process of corresponding member are as follows:
Filter out all members for not buying brand F.
Preferential action message flexibly defines according to actual needs, filters out the high meeting of matching degree according to default basic data
Member lays the foundation for the accurate distribution of subsequent favor information.
Optionally, by the member filtered out by the active policy (corresponding ordering strategy) in the preferential action message of correspondence into
Row sequence, different sequence sides can be corresponded to by obtaining active policy different in the corresponding degree of correlation list of the preferential action message
Formula, to be arranged for the purpose of the accurate distribution for improving favor information.It gives some instances below for reference:
The first, the people's certificate degree of correlation and price zone difference of each member that calculating sifting goes out；Wherein, people's certificate degree of correlation is root
The case where commodity corresponding according to member's history purchase type of offer, gives a mark to obtain；Price zone difference is reference price band and category valence
Difference between lattice band；It is ranked up according to people's certificate degree of correlation and price zone difference, obtains the corresponding correlation of preferential action message
Spend list.
Such as:
Type of offer is brand A, and active policy is brand consumption escalation policy, obtains the process of degree of correlation list are as follows:
1, history of the member that calculating sifting goes out at brand A buys situation, and the price of cumulative consumption is the score value for being people's certificate relationship, with
And price zone difference, such as: the category price zone of a member is 4 grades, and reference price band is 3 grades, and price zone difference is 1 grade；Separately
The category price zone of one member is 5 grades, and price zone difference is 2 grades；2, it is arranged from high to low or from low to high according to people's certificate degree of correlation
Sequence, when people's certificate correlation bottom is identical, second reference factor of the price zone difference as sequence；Such as: by people's certificate degree of correlation height from
Height arranges on earth, if they are the same, before the small row of price zone difference.
Second, according to default basic data, obtain member's loyalty of each member filtered out；According to member's loyalty
It is ranked up, obtains the corresponding degree of correlation list of preferential action message.
Such as:
Active policy is whole audience solicit patrons Flow Policy or the whole audience mentions objective single strategy or category solicit patrons Flow Policy or category mentions objective list
Strategy obtains the process of degree of correlation list are as follows: and 1, obtain the member's loyalty of each member screened, it can extract member's mark
Label, can also be realized with scoring mechanism；2, it sorts from high to low or from low to high by member's loyalty, it is corresponding to obtain preferential action message
Degree of correlation list.
The third, the member filtered out is ranked up according to people's certificate degree of correlation, obtains the corresponding phase of preferential action message
Pass degree list.
Such as:
Active policy is brand feedback strategy or single-item feedback strategy, obtains the process of degree of correlation list are as follows: will screen
Member out is ranked up from high to low or from low to high according to people's certificate degree of correlation, obtains the corresponding degree of correlation of preferential action message
List.
4th kind, the weighted scoring of membership buying and type of offer associated articles that calculating sifting goes out；The meeting that will be filtered out
Member is ranked up by weighted scoring, obtains the corresponding degree of correlation list of preferential action message.
Such as:
Type of offer is brand B, and active policy is that brand draws new strategy, obtains the process of degree of correlation list are as follows: 1, meter
Calculate the weighted scoring of membership buying commodity associated with brand B (commodity of associated brand are preset according to demand), weighting
Cumulative consumption amount, can be divided different gears by the regular flexible setting of scoring, and different stalls have different weight computings；2,
It after calculating, sorts from high to low or from low to high by weighted scoring, obtains the corresponding degree of correlation list of preferential action message.
Type of offer is single-item C, and active policy is that single-item draws new strategy, obtains the process of degree of correlation list are as follows: 1, meter
Calculating membership buying commodity associated with single-item C, (associated commodity are preset according to demand, and brand association can be used, improve
The reference information of a variety of commodity) weighted scoring, the regular flexible setting of weighted scoring can be by the commodity of different associated brands
Different weight computings is set to calculate cumulative consumption volume, as weighted scoring；2, after calculating, by weighted scoring from height to
It is low or sort from low to high, obtain the corresponding degree of correlation list of preferential action message.
It should be noted that screening corresponding member, sequence obtains degree of correlation list etc. by the way that different preferential classes is arranged
Type and associated active policy obtain different as a result, realizing different marketing purposes.The mode of sequence and screening can meet
It is freely combined in the case where logic, with no restriction.
Optionally, according to member's feature in default basic data and each movable membership information, matching obtains each movable meeting
After the corresponding member of member's information further include: when the corresponding member's quantity of an activity membership information is less than the activity membership information
When preferential quantity, member is expanded using crowd's broadcast algorithm, reaches the quantity of the corresponding member of the activity membership information preferential
Quantity.
Specifically, details are not described herein for the explanation of feature same with the above-mentioned embodiment in the present embodiment, refer to above-mentioned
Two embodiments.
Corresponding member is filtered out by preferential action message and movable membership information, obtains corresponding degree of correlation column
After table, two results are comprehensively considered according to allocation strategy, obtain final allocation result.
Default priority is arranged according to actual needs, such as: the corresponding type of offer of movable membership information > preferential activity letter
The brand type of offer in category type of offer > preferential action message in whole audience type of offer > preferential action message in breath >
Single-item type of offer in preferential action message.
Preset quantity in allocation strategy refers to the quantity for the favor information that each member should be assigned to, and can flexibly match
It sets, such as: 3 or 5 etc..
Optionally, further includes: when the quantity for the favor information being assigned there are a member is not up to preset quantity, automatically
For the remaining favor information of member's polishing.
Such as: preset quantity 2, totally 3 members, 2 preferential action messages (preferential quantity is respectively 2), 2 movable meetings
Member's information (preferential quantity is respectively 1).
The degree of correlation list of first preferential action message is from high to low are as follows: member 1, and member 2；
The degree of correlation list of second preferential action message is from high to low are as follows: member 3, and member 2；
The corresponding member of first activity membership information are as follows: member 1；
The corresponding member of second activity membership information are as follows: member 2；
Pre-set priority are as follows: first preferential action message of movable movable membership information > the second of membership information > the second
> the first preferential action message.
The favor information that member 1 is assigned to comes from: first movable membership information and first preferential action message.
The favor information that member 2 is assigned to comes from: second movable membership information and second preferential action message.
The favor information that member 3 is assigned to comes from: second preferential action message, automatically will be from remaining the because being discontented with 2
One preferential action message polishing.
In the present embodiment, preferential action message and movable membership information are used in combination and distribute favor information for member, together
When meet the marketing purpose of multi-angle (multiple brands, category while promoting), by big data technology member (i.e. customer) not
Suitable favor information can be matched for its amount body by its history consumer behavior before consumption, promote its consumption；And distribution is preferential
Information is obtained according to default basic data multi dimensional analysis, and the subjective factor influence of marketing plan personnel is lower, and matching is accurate
Degree is high, substantially increases the usage experience of member.
In a system embodiment of the invention, as shown in figure 5, a kind of personalized distribution system of favor information, packet
It includes:
Module 10 is obtained, for obtaining movable membership information；
Matching module 50, for according to the member's feature preset in basic data and movable membership information, matching is obtained pair
The member answered；
Distribution module 40 is used for corresponding (type of offer the is corresponding) favor information (favor information of movable membership information
It presents in different forms, such as: discount coupon, red packet information, bonus information etc.) distribute to and (be matched to) member.
Specifically, can be corresponding by member's feature of delineation member when wanting to distribute favor information for particular member
Member distributes favor information.
Movable membership information includes: type of offer, preferential quantity and member's feature, these three all can be clever according to actual needs
Setting living.
Type of offer includes: 1) whole audience favor information, 2) category favor information, 3) brand favor information and 4) single-item is preferential
Information etc..
Preferential quantity refers to the quantity of favor information, such as: the whole audience expires 500-230 and there was only 3000.
Default basic data be to data source (such as: retailer do shopping receipt transaction data, information of goods information data, member
Information data and shops's information data) import (such as: uncorrected data load, data cleansing, data accuracy inspection, data map,
Standardization number) after, corresponding label is stamped for each member according to these data, and, the sale sorted out summarizes data,
It is used for the subsequent corresponding member of screening.
Each member has multiple member's labels, defines according to actual needs.Sale summarizes data (example according to actual needs
Such as: the data that member's feature of definition needs) determine the data form of specific dimension.
Specific example refers to corresponding embodiment of the method, and details are not described herein.
Optionally, as shown in Figure 6, further includes: the quantity of extension module 60, the member for obtaining when matching is less than activity
When preferential quantity in membership information, member is expanded using crowd's broadcast algorithm, makes the quantity arrival activity membership information of member
In preferential quantity.
Specifically, crowd's broadcast algorithm can be used: Look-alike algorithm, certainly, other achievable target members extensions
Algorithm be also suitable, and with no restriction.
Crowd's broadcast algorithm is to find from remaining member and the assigned the most similar member of member, fortune
Logic when row includes:
1, consider feature: the score of all labels for the member being assigned, type of offer corresponding brand and category.Product
The score rule of board and category can be defined flexibly, such as: the accumulative gold of this brand, category is bought in each assigned member 52 weeks
Volume is as score.
2, it calculates and is assigned population centers' point, such as: the average value of all features.
3, member and assigned people of each preference shops within the scope of the corresponding movable shops of movable membership information are calculated
The distance of group center's point.
4, by widened member needed for apart from ascending sequence, selecting.
In the present embodiment, favor information, point of this favor information are distributed to it by delineation member's characteristic matching member
It is suitble to marketing demand to use when having clear member group with mode, accurately distributes to particular member, improves the matching of each member
Precision.
In another system embodiment of the invention, as shown in fig. 7, a kind of personalized distribution system of favor information,
Include:
Module 10 is obtained, for obtaining allocation strategy, at least one preferential action message and at least one movable member's letter
Breath.
Screening module 20, for filtering out member corresponding with each preferential action message respectively according to basic data is preset.
Sorting module 30, for respectively will the corresponding member of each preferential action message by the corresponding activity of preferential action message
Tactful (corresponding ordering strategy) is ranked up, and obtains the corresponding degree of correlation list of each preferential action message；Active policy is different,
The ordering strategy of each member can be different, also different in degree of correlation list, determine according to the actual situation.
Matching module 50, for according to the member's feature preset in basic data and each movable membership information, matching to be obtained
The corresponding member of each activity membership information.
Distribution module 40, for corresponding related according to each movable corresponding member of membership information and each preferential action message
List is spent, is to be preset in each member distribution allocation strategy preset in basic data according to the pre-set priority in allocation strategy
The favor information (favor information is presented in different forms, such as: discount coupon, red packet information, bonus information etc.) of quantity.
Specifically, preferential action message includes: type of offer, preferential quantity, active policy.Type of offer, preferential quantity
With active policy flexible setting according to actual needs.The definition of type of offer and preferential quantity is identical as the above system embodiment,
Details are not described herein.
There are many active policies, such as: brand feedback strategy, single-item consumption escalation policy, the whole audience mention objective single strategy, brand
New strategy, category solicit patrons Flow Policy, whole audience solicit patrons Flow Policy etc. are drawn, it is associated with type of offer.
In preferential action message, according to different type of offer, different active policies is set, improves sale to reach
The different marketing purposes such as amount, raising member's amount.
Default basic data is identical as the above system embodiment, and details are not described herein, it should be noted that the present embodiment is more
By the way of the distribution of preferential action message, therefore, calculates sale and summarizes data other than considering member's feature, it is also necessary to examine
Consider the needs of preferential action message.
The member that each preferential action message is directed to is different, and screening process naturally also has difference, according to preferential activity
Active policy screening in information.
Optionally, screening module 20 filter out corresponding with a preferential action message respectively according to default basic data
Member includes various ways, is determined according to the active policy of actual set, names some examples for reference:
The first, screening module 20 is calculated in default basic data in preset time period according to default basic data
(such as: 52 weeks) the monovalent specific quantile of each member visitor；Specific quantile and the active policy phase in preferential action message
It is corresponding；And according to specific quantile and the preferential rear amount of money, corresponding member is filtered out according to active policy；Wherein, after preferential
The amount of money is calculated according to the type of offer in preferential action message.
Second, screening module 20 calculates the corresponding reference of type of offer according to the type of offer in preferential action message
Price zone；And calculate the corresponding category price zone of type of offer of each member in default basic data；And according to ginseng
Price zone and the category price zone of each member are examined, filters out corresponding member according to the active policy in preferential action message.
The third, screening module 20 calculates all according to the type of offer in default basic data and preferential action message
People's certificate degree of correlation of member；And according to the active policy in people's certificate degree of correlation and preferential action message, filter out corresponding meeting
Member.
4th kind, screening module 20 filters out not according to the active policy in default basic data and preferential action message
Bought the member of the corresponding commodity of type of offer in preferential action message.
Preferential action message flexibly defines according to actual needs, filters out the high meeting of matching degree according to default basic data
Member lays the foundation for the accurate distribution of subsequent favor information.
Optionally, sorting module 30, the member filtered out is (corresponding by the active policy in the preferential action message of correspondence
Ordering strategy) it is ranked up, difference can be corresponded to by obtaining active policy different in the corresponding degree of correlation list of preferential action message
Sortord, by improve favor information it is accurate distribution for the purpose of be arranged.It gives some instances below for reference:
The first, sorting module 30 obtains member's loyalty of each member filtered out according to default basic data；With
And be ranked up according to member's loyalty, obtain the corresponding degree of correlation list of preferential action message.
Second, the member filtered out is ranked up by sorting module 30 according to people's certificate degree of correlation, obtains preferential activity letter
Cease corresponding degree of correlation list.
The third, sorting module 30, the people's certificate degree of correlation and price zone difference of each member that calculating sifting goes out；Wherein, people
The case where certificate degree of correlation is commodity corresponding according to member's history purchase type of offer gives a mark to obtain；Price zone difference is reference price
Difference between lattice band and category price zone；And be ranked up according to people's certificate degree of correlation and price zone difference, obtain preferential work
The dynamic corresponding degree of correlation list of information.
4th kind, sorting module 30, the weighted scoring of membership buying and type of offer associated articles that calculating sifting goes out；With
And be ranked up the member filtered out by weighted scoring, obtain the corresponding degree of correlation list of preferential action message.
It should be noted that screening corresponding member, sequence obtains degree of correlation list etc. by the way that different preferential classes is arranged
Type and associated active policy obtain different as a result, realizing different marketing purposes.The mode of sequence and screening is patrolled meeting
It is freely combined in the case where volume, with no restriction.
Specific example in the present embodiment refers to corresponding embodiment of the method, and details are not described herein.
Optionally, further includes: extension module 60, for being less than the activity when the corresponding member's quantity of an activity membership information
When the preferential quantity of membership information, target members are expanded using crowd's broadcast algorithm, make the quantity arrival activity meeting of target members
Preferential quantity in member's information.
Specifically, details are not described herein with the explanation of same characteristic features in the above system embodiment in the present embodiment, refer to
Above-mentioned two system embodiment.
Corresponding member is filtered out by preferential action message and movable membership information, obtains corresponding degree of correlation column
After table, two results are comprehensively considered according to allocation strategy, obtain final allocation result.
Default priority is arranged according to actual needs, such as: the corresponding type of offer of movable membership information > preferential activity letter
The brand type of offer in category type of offer > preferential action message in whole audience type of offer > preferential action message in breath >
Single-item type of offer in preferential action message.
Preset quantity in allocation strategy refers to the quantity for the favor information that each member should be assigned to, and can flexibly match
It sets, such as: 3 or 5 etc..
Optionally, further includes: polishing module 70, for being not up to when the quantity for the favor information being assigned there are a member
It is automatically the remaining favor information of member's polishing when preset quantity.
Specific example refers to corresponding embodiment of the method, and details are not described herein.
In the present embodiment, preferential action message and movable membership information are used in combination and distribute favor information for member, together
When meet the marketing purpose of multi-angle (multiple brands, category while promoting), by big data technology member (i.e. customer) not
Suitable favor information can be matched for its amount body by its history consumer behavior before consumption, promote its consumption；And distribution is preferential
Information is obtained according to default basic data multi dimensional analysis, and the subjective factor influence of marketing plan personnel is lower, and matching is accurate
Degree is high, substantially increases the usage experience of member.
It should be noted that above-described embodiment can be freely combined as needed.The above is only of the invention preferred
Embodiment, it is noted that for those skilled in the art, in the premise for not departing from the principle of the invention
Under, several improvements and modifications can also be made, these modifications and embellishments should also be considered as the scope of protection of the present invention.
Claims (10)
1. a kind of personalized distribution method of favor information characterized by comprising
When getting movable membership information, according to member's feature in default basic data and the movable membership information,
With obtaining corresponding member；
The corresponding favor information of the activity membership information is distributed into the member.
2. the personalized distribution method of favor information as described in claim 1, which is characterized in that further include:
When the quantity for the member that matching obtains is less than the preferential quantity of the movable membership information, expanded using crowd's broadcast algorithm
The big member makes the quantity of the member reach the preferential quantity.
3. the personalized distribution method of favor information as described in claim 1, which is characterized in that further include:
When getting allocation strategy, at least one preferential action message and at least one movable membership information, according to default base
Plinth data filter out member corresponding with each preferential action message respectively；
The corresponding member of each preferential action message is ranked up by the active policy in the preferential action message respectively,
Obtain the corresponding degree of correlation list of each preferential action message；
According to member's feature in default basic data and each movable membership information, matching obtains each movable member's letter
Cease corresponding member；
According to the corresponding member of each movable membership information and the corresponding degree of correlation list of each preferential action message, according to
Pre-set priority in the allocation strategy is that each member in default basic data distributes present count in the allocation strategy
The favor information of amount.
4. the personalized distribution method of favor information as claimed in claim 3, which is characterized in that further include:
It is automatically member's polishing when the quantity for the favor information being assigned there are a member is not up to the preset quantity
The remaining favor information.
5. the personalized distribution method of favor information as claimed in claim 3, which is characterized in that the default basis of the basis
Data, filtering out member corresponding with the preferential action message includes:
According to default basic data, specific point of each member visitor unit price in preset time period is calculated in default basic data
Digit；The specific quantile is corresponding with the active policy in the preferential action message；
According to the specific quantile and the preferential rear amount of money, corresponding member is filtered out according to the active policy；
Wherein, the preferential rear amount of money is calculated according to the type of offer in the preferential action message.
6. a kind of personalized distribution system of favor information characterized by comprising
Obtain module, acquisition activity membership information；
Matching module, for according to the member's feature preset in basic data and the movable membership information, matching to be corresponded to
Member；
Distribution module, for the corresponding favor information of the activity membership information to be distributed to the member.
7. the personalized distribution system of favor information as claimed in claim 6, which is characterized in that further include:
Extension module when preferential quantity for being less than in the movable membership information when the quantity of member that matching obtains, is adopted
Employment group diffusion algorithm expands the target members, and the quantity of the member is made to reach the preferential number in the movable membership information
Amount.
8. the personalized distribution system of favor information as claimed in claim 6, it is characterised in that:
The acquisition module is further used for obtaining allocation strategy, at least one preferential action message and at least one movable meeting
Member's information；
The personalized distribution system of the favor information, further includes:
Screening module, for according to basic data is preset, filtering out member corresponding with each preferential action message respectively；
Sorting module, for respectively by the corresponding member of each preferential action message by the activity in the preferential action message
Strategy is ranked up, and obtains the corresponding degree of correlation list of each preferential action message；
The matching module is further used for according to the member's feature preset in basic data and each movable membership information,
Matching obtains the corresponding member of each activity membership information；
The distribution module is further used for according to each movable corresponding member of membership information and each preferential activity letter
Corresponding degree of correlation list is ceased, is each member point in default basic data according to the pre-set priority in the allocation strategy
Favor information with preset quantity in the allocation strategy.
9. the personalized distribution system of favor information as claimed in claim 8, which is characterized in that further include:
Polishing module, for when the quantity for the favor information being assigned there are a member is not up to the preset quantity, automatically
For the remaining favor information of member's polishing.
10. the personalized distribution system of favor information as claimed in claim 8, which is characterized in that the screening module is used for
According to default basic data, filtering out member corresponding with the preferential action message includes:
It is single that each member visitor in preset time period is calculated in default basic data according to default basic data in screening module
The specific quantile of valence；The specific quantile is corresponding with the active policy in the preferential action message；
And according to the specific quantile and the preferential rear amount of money, corresponding member is filtered out according to the active policy；
Wherein, the preferential rear amount of money is calculated according to the type of offer in the preferential action message.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201910023753.7A CN109784996A (en) | 2019-01-10 | 2019-01-10 | A kind of personalized distribution method and system of favor information |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201910023753.7A CN109784996A (en) | 2019-01-10 | 2019-01-10 | A kind of personalized distribution method and system of favor information |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN109784996A true CN109784996A (en) | 2019-05-21 |
Family
ID=66499356
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201910023753.7A Pending CN109784996A (en) | 2019-01-10 | 2019-01-10 | A kind of personalized distribution method and system of favor information |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN109784996A (en) |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN113554385A (en) * | 2021-05-27 | 2021-10-26 | 北京每日优鲜电子商务有限公司 | Distribution robot control method, distribution robot control device, electronic equipment and computer readable medium |
Citations (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN107862549A (en) * | 2017-11-03 | 2018-03-30 | 北京翰宁智能科技有限责任公司 | A kind of reward voucher put-on method and system |
-
2019
- 2019-01-10 CN CN201910023753.7A patent/CN109784996A/en active Pending
Patent Citations (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN107862549A (en) * | 2017-11-03 | 2018-03-30 | 北京翰宁智能科技有限责任公司 | A kind of reward voucher put-on method and system |
Non-Patent Citations (1)
| Title |
|---|
| 智慧店铺夏夏: "智慧店铺微信发券增加会员属性筛选", 《HTTPS://WWW.JIANSHU.COM/P/CA928DC2797B》 * |
Cited By (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN113554385A (en) * | 2021-05-27 | 2021-10-26 | 北京每日优鲜电子商务有限公司 | Distribution robot control method, distribution robot control device, electronic equipment and computer readable medium |
| CN113554385B (en) * | 2021-05-27 | 2024-01-05 | 广东中顺信息科技有限公司 | Distribution robot control method, distribution robot control device, electronic equipment and computer readable medium |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN109584018A (en) | A kind of internet marketing cloud service platform | |
| Ramanathan | Supply chain collaboration for improved forecast accuracy of promotional sales | |
| US20040088221A1 (en) | System and method for computing measures of retailer loyalty | |
| US20130066740A1 (en) | Commerce System and Method of Controlling the Commerce System Using Personalized Shopping List and Trip Planner | |
| WO2016088109A1 (en) | System and methods for evaluating and increasing customer engagement | |
| WO2003081376A2 (en) | Targeted incentives based upon predicted behavior | |
| CN106875227A (en) | A kind of Xian Xia shops rebate system based on IOS and Android | |
| Kumar et al. | The impact of psychological pricing strategy on consumers' buying behaviour: a qualitative study | |
| CN109711906A (en) | A kind of distribution method and system of favor information | |
| Tauras et al. | The role of retail prices and promotions in determining cigarette brand market shares | |
| Cotterill | Food retailing: Mergers, leveraged buyouts, and performance | |
| CN109784996A (en) | A kind of personalized distribution method and system of favor information | |
| Wright | Slotting contracts and consumer welfare | |
| Ling | Customer Relationship Management: Case study Coca-Cola Company | |
| CN106485550A (en) | accounting method and terminal | |
| CN112446746A (en) | Consumption rebate management platform system and rebate method | |
| Ali et al. | Role of Social Media in Retail Network Operations and Marketing to Enhance Satisfaction: An Analytical Approach | |
| McClelland | Pricing for profit in retailing | |
| Aneja | Marketing and selling in India in the year 2000 | |
| Suthar et al. | Factors triggering poverty in District Tharparkar Sindh, Pakistan | |
| Noriega-Goodwin | Doubled SNAP Dollars and Nudges: An Analysis of Two Pilot Programs Aimed at Increasing the Purchase of Healthy Foods | |
| Gallo et al. | Couponing's growth in food marketing | |
| Meloy et al. | Consumer Segmentation Analysis of Grocery Coupon Users | |
| CN106022559A (en) | Data display method and device | |
| Jham et al. | Domino's Pizza India Ltd.: Driving Business Growth through Consumer Engagement |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| RJ01 | Rejection of invention patent application after publication | ||
| RJ01 | Rejection of invention patent application after publication |
Application publication date: 20190521 |